An Algorithmic Embedding of Graphs via Perfect Matchings

  • Authors:
  • Vojtech Rödl;Andrzej Rucinski;Michelle Wagner

  • Affiliations:
  • -;-;-

  • Venue:
  • RANDOM '98 Proceedings of the Second International Workshop on Randomization and Approximation Techniques in Computer Science
  • Year:
  • 1998

Quantified Score

Hi-index 0.00

Visualization

Abstract

Recently Komlós, Sárközy, and Szemerédi proved a striking result called the blow-up lemma that, loosely speaking, enables one to embed any bounded degree graph H as a spanning subgraph of an Ɛ-regular graph G. The first proof given by Komlós, Sárközy, and Szemerédi was based on a probabilistic argument [8]. Subsequently, they derandomized their approach to provide an algorithmic embedding in [9]. In this paper we give a different proof of the algorithmic version of the blow-up lemma. Our approach is based on a derandomization of a probabilistic proof of the blow-up lemma given in [13]. The derandomization utilizes the Erdös-Selfridge method of conditional probabilities and the technique of pessimistic estimators.